This thesis examines the following questions about the behaviour of the US Treasury bill market, and published forecasts of US Treasury bill yields: - Are expert forecasts of Treasury bill yields, as revealed in survey data, more accurate than simple alternatives? - Can they be used to make money by trading Treasury bill futures, and does this imply that the futures market is inefficient? - Is there information in the survey data relevant to predicting the volatility of Treasury bill yields? - How does this compare to more conventional volatility forecasts, such as those from the implied volatility in three months treasury bills futures-options; historical volatility, and time-series based volatility forecasts based on the popular GARCH (1,1) model - Can these forecasts be used to make money by trading options on Treasury bill futures?